Calibrating An Inductance Standard JAMES E . WACHTER, P r o j e c t E n g i n e e r

Definition of Inductance Standard

Ask an engineer for the definition of an inductance standard and he will prob- ably tell you that it is a coil or inductor having an accurately known, highly stable inductance. This is the definition implied. However, a better definition would be, an inductor having highly stable and accurately determined param- eters; i.e., inductance (L) , distributed capacitance (cd) , and resistance ( R ) .

In such a standard, the parameters L and cd are readily measured but the accurate measurement of R is extremely

L’ difficult. This is true because, in general, the more useful coils have a relatively high Q and consequently a very small value of R. R is so small in fact, that it is often swamped by the losses in any measuring equipment used, and there- fore is very difficult to isolate. Since Q

is a function of R (Q =-), it follows

that if Q and L can be determined, the value of R is firmly fixed.

Methods for Measuring Q

The problem now is to measure Q. An investigation into the possible methods of measuring Q has been made and the following conclusions drawn: 1. The frequency variation method in-

volves the ratio of the frequency at resonance to the difference in fre- quency between the two half-power points on the Q-versus-frequency curve. This method was found un- suitable because of the variation of

WL R

,

L/

YOU WILL FIND . . . A Linear Detector for FM Deviation Measurements ..................... 5

BRC FILM GAUGE USED ro MEASURE AlRCRAFT ORGANlC FlNlSH THICKNESS 7

Editor‘s Note ....................... 8

2.

3

4

Figure 1 . The author shown determining C at a V, /V ratio, using the modified and specially calibrated Q-Meter.

impedance of the Q circuit with fre- quency and the fact that the resonant frequency is different from the fre- quency for maximum voltage. Injecting an AM signal into the Q circuit and measuring the attenuation of the side bands was rejected be- cause errors in the amplitude of the side bands are caused both by cou- pling to the Q circuit and any asym- metry of modulation. Injecting an AM signal into the Q circuit and measuring the phase angle of the detected signal was also re- jected due to the error introduced in coupling to the Q circuit and the difficulty in accurately measuring the phase angle. Q as determined from measurements on a BRC G-Meter, Type 192-A, is quite accurate and not difficult to determine, but because the G-Meter provides only two measurement fre- quencies, 1 mc and 30 mc, this means

was found unsuitable as a general method.

5 . A variation of the “Q by C” method was found to be most suitable, be- cause the frequency remains fixed for these measurements, and the effects on the Q circuit due to varying fre- quency are eliminated. Also, the measurement requires basically an injection system, a Q circuit with a variable capacitor, and an oscillator, all of which are available in a Q- Meter.

Q Defined ic Up to this point, the term “Q’ has been used rather loosely. To aid in this discussion, it might be well to define here the several terms of Q with which we will be dealing:

W L Q - true Q of the inductor; i.e., - R Q,-effective Q; Le., the Q of the

inductor mounted on a Q meter, exclusive of all Q-Meter losses.

B O O N T O N R A D I O C O R P O R A T I O N

T H E BRC NOTEBOOK is publzshed four zzmes a year by the Boonton Rad80 Corporatzon. It I S mazled free of charge t o screntrsts, engzneers and other $mer- estcd persons zn the comlltuntcat*ons and electronzcs fzelds The contents may be reprrnted only wrth wrztten permrs- saon f r o m d e edztor. Your comments and suggest tons are welcbme, and should be addressed i o ' Edztor, T H E BRC N O T E B O O K , Boonton Radto Corporation, Boonton, N . J .

Q, -circuit Q; Le., the Q of the Q- Meter resonant circuit, including rhe inductor.

Q,-indicated Q; Le., the Q of the Q-Meter resonant circuit as in- dicated by the meter. This value includes the calibration errors of the Q-Merer.

Determination of b

From the preceding definitions, it is seen that QI,, the Q that the inductor appears to have when associated with additional circuitry, is the most useful value. If necessary, true Q can be de- rived from this value. To measure Qe, a substitution method is used whereby the conductance of the Q circuit is deter- mined first with a well shielded high-Q coil and again with the same high-Q coil plus the inductor being evaluated. The difference between the two determina- tions is the conductance of the unknown. This can be shown mathematically. Referring to the voltage-versus-capaci- tance curve (Figuce 2 ) of the Q-Meter tank circuit, the conductance of the Q circuit with the high-Q coil can be expressed as:

The conductance with the two coils is then:

If the ratio 3 is made equal to v1

3 then the conductance of the V-'

unknown is:

With the conducrance of the inductor known, it is a simple step to compute the effective Q:

It should be noted that the inductor should have a Q of more than 100, since the foregoing equations have been de- rived using this assumption.

While the preceding analysis is straightforward, the actual measure- ments are involved. From equations ( 3 ) and ( 4 ) it is seen that because w is readily determined to a high degree of accuracy using 3 crystal frequency calibrator or frequency counter, the overall accuracy is dependent upon the determination of C,,, AC, and VJV and the ability to repeat specific -ratios of v,,/v.

Figure 2. Voltage versus capacirance curve of a Q-Meter tank circuit.

To make the measurements we have used a modified and specially calibrated Q-Meter ( Figure 1 ) . Additional bind- ing posts have been added to permit mounting two coils; one in the normal manner and the other from the HI post to ground. A means has been provided whereby the QMeter B+ voltage is externally regulated and monitored. Direct connection may be made to the cathode of the Q-voltmerer tube. In ad- dition, a calibrated high-ratio gear drive is used to operate the main Q-capacitor and a parallel group of three micrometer type vernier capacitors, having a total range of abour 100ppf, replaces the usual vernier. This permits the main tuning dial to remain in a fixed position while a wide range of A C readings is made.

vol tage Katio

Setting up the V,,/V ratios neces- d sitates the L I S ~ of two specialized pieces of eqLiipment; an instrument to provide precise levels of a 1000 cps signal ( voltage calibrator) and an instrument to provide several very stable DC volt- ages (bucking voltage source). The 1000 cps Source is used to calibrate the DC Source in the following manner: The 1000 cps source is connected to the capacitor terminals of the Q-Meter, whose oscillator is made inoperative by setting between ranges. The DC source is connected through a microammeter to the cathode of the Q-voltmeter tube (See figure 3 ) . Q-Meter zero is ac- curately set and periodically checked. The Q-Meter B+ supply voltage is also checked periodically to insure stable conditions. The level of the 1000 cps signal is adjusted to give il reading well LIP the Q scale of the Q-voltmeter. This is the resonance reference v,, = 1.

The DC source is then adjusted, in the 1 position, to give a zero reading on the

'microammeter in the lead to the Q- voltmeter tube, indicating that the volt- age is of the same level as the voltage delivered to the Q-voltmerer. The 1000 cps source is now set to give exactly 0.9 of the previous signal and the DC source is switched to the next output and ad- justed for zero meter reading. This is repeated for several levels; i.e., 0.8, 0.7 0.6.

It can be seen that the DC source is now a memory for the various voltages appearing across the Q-Meter tuning capacitor, enabling the user to set and reset precisely to any desired voltage.

With the 1000 cps source removed and the Q-Meter oscillator set to the desired test frequency, the inductor (or inductors) to be measured is connected and the Q-Meter is resonated with the main and vernier capacitors. Capacitor settings are recorded and correspond to the resonating capacitance Co. The DC source is set to the 1 reference position and the Q-Meter XQ level is adjusted to give a zero reading on the micro- ammeter. The DC source is then switched to rhe 0.9 reference position and the Q-Meter is detuned on either side of resonance, using the vernier ca- pacitors. The difference between the capacitor settings, ar the point on either side of resonance where the current meter reads zero, is the AC value for the ratio V,,/V = 0.9. These capacitor ii settings are also recorded, and the pro- cedure is repeated for the VJV ratios of 0.8, 0.7, etc.

W

2

T H E N O T E B O O K

Deternainiizg Cdpncitame - All that remains is to accurately deter- mine the capacitance at the recorded capacitor settings before applying equa- tions ( 3 ) and ( 4 ) .

Modifications made to the Q-capaci- tors permit settings to be repeated to a very fine degree. This is necessary, because it is required that the Q-Meter be turned off to calibrate the capacitor.

The actual capacitance is measured by connecting a sensitive capacitance bridge to the capacitance terminals of the Q- Meter whose capacitors are set to a previously recorded value. The bridge is then balanced using a precision ca- pacitor. All known corrections to the precision capacitor are applied and cor- rection for the leads from the bridge to the Q-Meter is made.

The preceding is sufficient for the difference in capacitance (AC) data, but for absolute capacitance (C(,) data; additional corrections are necessary. A correction for the Q-voltmeter level is required, because the Q capacitance is measured with the Q-Meter turned off. This correction is determined through the use of a second Q-Meter (No. 2 ) . The oscillator of the Q-Meter to be checked is disabled (set between ranges) and the capacitance terminals of this Q- Meter are connected to those of Q-Meter No. 2. Number 2 Q-Meter has a coil connected in its tank circuit and its oscillator is operated at the test fre- quency. The coil is selected so that some low value of capacitance is required to resonate with it. If the capacitance re- quired is 80ppf, then about 40ppf will be supplied by the Q-capacitor of each Q-Meter.

The Q-Meter under test is turned off and its vernier capacitors are used to resonate the tank circuit of Q-Meter No. 2; then it is turned on again, and the process is repeated. The difference in the two settings of the vernier is the capacitance attributable to the Q-volt- meter a t the particular voltage (or Q) level. Different voltage levels can be selected by changing the setting of rhe XQ control on Q-Meter No. 2 and the capacitance correction versus Q level can be plotted as shown in Figure 4.

A final correction to the capacitance values may be necessary. When the Q- capacitor of a Q-Meter is calibrated, some small capacitance existing between the Q-Meter HI terminal and the cabinet (ground) is included in the calibration. A shielded coil connected to the coil terminals of the Q-Meter (shield con- n’ected to the LO terminal), causes some

w

I I Q METER TANK * VOLTMETER

Figure 3. Connections used for setting up V , / V rrrtios.

~~~ SOURCE

- - VOLTAGE CALIBRATOR

of the capacitance to shift from HI terminal to cabinet, to HI terminal to shield. This is called the proximity effect and is exceedingly difficult to define. In all but the most rigorous determinations, this effect may be neg- lected without seriously affecting the accuracy of the result. Without develop- ing a lengthy and involved procedure for determining this effect, it may be said that generally any shielded coil hav- ing overall dimensions similar to the shielded coils manufactured by Boonton Radio Corporation ( 3-inch diameter shield Fans) will produce a proximity effect of approximately -0.4ppf when mounted on a Q-Meter with the coil base about 1 inch above the Q-Meter top panel. The figure will decrease with a decrease in the shield can diameter or an increase in the distance between the shield can and the Q-Meter.

Recapitulating the capacitance cor- rections:

Ci - Capacitance indicated by Q- Meter Q-capacitor.

I C , . - Correction indicated by pre- cision capacitor.

r+C, -Correction for leads from capacitance bridge to Q- Meter.

t C , - Correction for Q-voltmeter level.

-C,) - Correction for proximity effect.

It should be noted that for best results all measurements should be conducted in a temperature and humidity con- trolled atmosphere so that both the in- ductor under test and the measuring equipment are not affected by these conditions.

The AC values derived using the de- scribed procedures are used in equation ( 3 ) and a value of G, is obtained for each V,,/V ratio. An indication of the care and accuracy of measurement is ap- parent by the degree of coincidence of the several values of G, for each test frequency The C,, value measured at each test frequency is used in Conjunc- tion with the average G, value for that frequency in equation ( 4 ) to yield the

effective Q of the inductor True inductance and distributed ca-

pacitance of the inducror can be found by applying data obtained from the previous measurements to the following equations:

/ 1 1 \ \ - - -p j

( 5 ) L= 4 7T2 (C,,, - c,,, )

Where: C,,, and C,,2 are the capacitances neces- sary to resonate the coil at frequencies f t and f, respectively, and n is the ratio of f 2 to fl.

Figure 4. curve for a Q-Meter.

Capacitance correction versus Q - l e v e l

All of the true and effective para- meters may now be determined by applying the following equations:

( 7 ) Effective inductance ( L e ) = L

1 - w”Ca (8 ) Effective resistance (R,) =

OL, R Qe

- (1 - O*LCd)2

OL Ca ( 9 ) T r u e Q = - z Q Q , ( l + - ) R c, -

3

B O O N T O N R A D I O C O R P O R A T I O N

Bibliography BRC Q-Standard, Type 51 3-A

A specific example of an inductor developed through use of the described methods is the BRC Q-Standard, Type 5 13-A. The nameplate information for this inductor includes L, Cd, Qe at three frequencies, and another term, Qi, at the same three frequencies. In this case, Qi is the Q that would be indicated by a Q-Meter having average loss and zero calibration error. The Qi information was derived through an analysis of pro- duction indicated Q checks made on Q- Meters manufactured by BRC and relat- ing the result to the measured Qe value. Of course, in production the procedures just outlined would be impractical, therefore, a comparison method was de- veloped. At each of the three frequen- cies involved, the production coil is compared to a standard, which has been established using the rigorous method described above. A precisely calibrated Q-Meter is used for the comparison, although its .accuracy has only higher order effects on the results.

Suppose that at one of the frequencies, the difference between the production coil and the standard, as compared on a Q-Meter, is given by AC and AQ. The functions of standard and production coils would then be defined as follows:

Coil Qix

Qex

Where:

( IO) Qis = Qi + AQ (11) c,,, = C” + AC (12) Qrs=

Using the same process to obtain differ- ence data at the other two frequencies, (2,.s can be found at each frequency. Distributed capacitance and inductance of the unknown are given by:

(14) L,=

Where the subscripts 1 and 2 refer to measurements at 2 frequencies.

To reduce the possibility of errors in manipulation, equations ( 12 ) , ( 13 ) , and (14) have been transformed to nomograms for use by production per- sonnel ( See Figure 5 ) .

Figure 5. BRC inspector shown using a nomo- gram to determine the effective Q of (I Type 513-A inducfor.

A great deal of care has been taken in the physical design and manufacture of the Q-Standards to insure their stability. The coil form is mounted on a copper base, which is fitted to a shield can. The unit is hermetically sealed, evacuated, and filled with an inert gas to a pressure of 1 psi above atmospheric pressure. Leads are brought through the base to banana plug connectors, which may be replaced without breaking the seal. The high potential connector is isolated from the base by a low-loss ceramic seal.

The care taken in determination and production of the Q-Standards is attested to by the fact that not a single coil has been returned for mechanical failure or deterioration of electrical specifications. In some cases, Q-Standard users involved in government work have been required by the cognizant government agency to have their Q-Standards periodically re- checked by BRC. In each such case it has been found that the nameplate informa- tion was well within the original spe- cification tolerance and no corrections of any kind were required.

d 1. Application Instructions for the Q-

Standard Type 5 13-A. 2. Hopf, W., Boonton Radio Corp.,

“Precision Measurement of Q by AC’, Feb. 1951.

3. Kang, C. L. and Wachter, J., “The 0-Standard”. BRC Notebook No. 1, $ring, 1954.

4. Kang, C. L., “Circuit Effects on Q’, BRC Notebook No. 8, Winter, 1956.

5 . “A Standard for Q and L”, BRC Notebook No. 12, Winter, 1957.

BRC manufactures two Q-Standards; the Type 513-A, discussed in the above article, and the Type 518-A. Each Type 513-A Q-Standard i s individually calibrated and marked with its true inductance (L) , distributed capacitance (Cd), effective Q(Q,), and indicated Q ( Q , ) at 0.5, 1.0, and 1.5 megacycles. Because these para- meters are accurately known and highly stable, this standard may be used for providing pre- cisely known supplementary Q-circuit inductance desireable for many impedance measurements by the parallel method, as well as a means for checking the Type 260-A and 160-A Q-Meters. The Type 518-A Q-Standard on the other hand i s a precision inductor designed primarily for use in checking the overall operating accuracy of Q-Meters Type 260-A and 160-A.

NOW AVAILABLE FOR IMPROVED CUSTOMER SERVICE

BELL SYSTEM

”1 E, .17- 1.1

TWX BOONTON NJ 866

Boonton Radio Corporation Boonton, N. J .

‘d

4

T H E N O T E B O O K

A linear Detector for FM Deviation Measurements F R A N K P . MONTESION, E d i t o r T h e N o t e b o o k

With the development of the Type 202 FM Signal Generators in 1946, there arose a need for a device for the precision measurement of frequency deviation. Such a device would be re- quired to provide laboratory accuracy and, at the same time, had to be simple and convenient for direct use on the production line. A detailed survey of available instrumentation revealed that such an instrument was not available and work was carried out at BRC on the development of this laboratory tool, con- current with the development of the FM Signal Generator.

The result of this project was an early prototype unit, which over the years has Endergone constant redesign and im- provement to become known as the BRC Type 208-A FM Linear Detector. These instruments have been in constant use in our standards laboratory and engi- neering and production departments during this period, and are currently used to calibrate the Types 202-E and 202-F, FM-AM Signal Generators.

At the request of several customers who had a need to perform similar measurement of frequency deviation, the Type 208-A FM Linear Detector has been put into production and is now available for sale.

‘1

Operating Principles

The basic circuit elements of the FM Linear Detector are shown in block

Detector are conventional circuits that operate to produce a signal usable for detection purposes. Actual detection is accomplished in the limiter and discrim- inator circuits shown in simplified schematic form in Figure 3. A type 6C4 triode, tuned over a frequency range of 27 to 54 megacycles is used as an RF oscillator. The output from the RF oscil- lator is fed to a Class C frequency- doubling stage tuned to the second har- monic of the oscillator frequency ( 54 to

ycles) which drives a Class rating as a frequency dou-

bling stage on the high frequency range

.

Figure 1. Type 208-A FM Linear Defector.

(108 to 216 megacycles). The RF out- put from the doubler-amplifier stages, together with the output of the FM signal generator under test, are fed into a mixer stage. The difference frequency produced by the mixer is then fed

stages of IF amplification to the limiter stage. After the first stage of IF amplification, a signal is fed through a cathode follower stage to the IF terminals for use in AM measure- ments.

petition rate or frequency of the incom- ing signal.

A low-pass filter is connected across the output of the discriminator to re-

red signal frequency com- to allow the instantaneous rise and fall with each dis-

charge of a current pulse. The demodu- lated output of the detector is a varying unidirectional potential directly propor- tional to the IF frequency.

Discriminator and Limiter

Referring to Figure 3, the input ei,, is an FM signal with a carrier frequency between 100 and 300 kc. The amplitude of this signal is sufficient to overswing the cutoff and zero bias limits of tube 807. During the part of the cycle when the tube is cut off, the plate potential will rise to the level of En. When the grid is positive, the plate current will

then that the minimum and maximum instantaneous plate potential 7 are held constant, producing wave which is squared off, top

and bottom, at definite fixed potentials and which is unaffected by possible variations in grid-swing amplitude. This square wave of plate voltage has a rep- etition rate equal to that of the input signal, eb,.

art of the cycle when the of tube 807 reaches its

peak value (ET%), capacitor C1 is charged through diode dl of tube 6H6 to a potential equal to El%. When the plate potential of tube 807 swings toward its

5

B O O N T O N R A D I O C O R P O R A T I O N

-- r---

1 F M I FOLLOWER DOUBLER ‘ EXTCRNAL I CATHODE

6 C 4 I 6AKS I GENERATOR

lowest value, capacitor C1 discharges through diode d:! of tube 6H6 in series with its load, Ri. This action causes one pulse of current to flow through resistor R + for each cycle of operation.

The total charge taken by capacitor C1, once each cycle, is CE. ( A small bias voltage in series with the charge diode dl effectively overcomes the contact potea- tial of both diodes dl and da, therefore, the effect of this potential may be dis- counted ) The portion of this total charge which passes through diode da and resistor R i of tube 6H6 i s equal to the total charge ( CErl) minus the resid- ual charge ( Ce,,llnn) .

With El< and ePllllll held constant, and the time constants of the charge and discharge circuits sufficiently small com- pared to the input wave period, the total quantity of current flowing through R,3 during each cycle is constant. An in- crease in the repetition rate (frequency) of the incoming signal will increase the number of current pulses per unit of time, thereby increasing the average value of current flowing through R3. Conversely, a decrease io the reperition rate of the incoming signal will decrease the average current through R3. The average potential across R.3 then is a perfectly linear function of the fre- quency of the incoming signal and the dynamic operation of the detector will result in essentially distortionless detec- tion.

I

Calibration

IF AMPLIFIER MIXER A MPLl Fl ER

DOUBLER -

The Linear Detector is accurately cali- brated at the factory to provide a voltage versus frequency deviation coefficient for frequency deviations up to k 3 0 0 kc at an intermediate frequency of 350 kc. However, it is advisable, because of com- ponent aging, to recalibrate the instru- ment periodically during normal use. Either of two methods, the Static Method or the Bessel Zero Method, may be used to accurately calibrate the Linear Detector.

/

LIMITER IF IF - AMPLIFIER - AMPLIFIER -

Static Method

6 A 6 4 6AK5

The Detector is interconneceed with an FM signal generator, a frequency calibrator, and an accurate DC measur- ing device. The generator is connected to the Detector’s RF INPUT terminals, the frequency calibrator is connected to the IF OUTPUT terminals, and the DC measuring device is connected to the

8 07 6AC7 6AC7 6AG7

-

qF OSCILLATOR c LOW-PASS FILTER ii

Figure 2. Basic circuit elements of the Type 208-A FM Linear Detector,

DEMOD OUT terminals. With the frequency dials on the signal

generator and the Detector set to the same frequency, the DC measuring de- vice will read zero. Advancing the De- tector frequency dial to provide 100 kc difference frequency, as indicated by the frequency calibrator, will cause a voltage reading to be indicated by the DC measuring device. This reading is noted and the Detector frequency dial is advanced again until a 200 kc difference frequency is indicated by che frequency calibrator. The voltage reading at the DEMOD OUT terminals is again noted. This procedure is repeated at each 100 kc increment until a 1 megacycle signal is indicated by the frequency calibcator, the output voltage being recorded for each step. The voltage readings obtained are the n plotted on a graph with voltage as the “Y” axis and frequency as the “X” axis. The resultant curve will yield a straight-line section, whose upper and lower limits indicate the frequency points in the frequency spectrum be-

DlSCRl MINATOR 7 tween which linear operation of the Detector should be expected. With the DC voltage for these two frequency de- viations known, any deviation (within the linear limits) may be ascertained by using the frequency versus voltage co’efficient.

Bessel Zero Method

This method of calibration requires the use of a signal generator, an accurate 10-kc audio signal source, and a hetero- dyne-type receiver containing a BFO. The RF output of the signal generator is connected to the receiver and the re- ceiver is tuned to the unmodulated carrier frequency of the generator to obtain a beat frequency of several hun- dred cycles, using the receiver’s BFO. This frequency is monitored with ear- phones or a voltmeter. With the signal generator modulation control set to pro- duce a 10-kc FM modulating signal, the FM deviation control is advanced slowly.

L2 R4 L l

‘J

Figure 3. Basic detector circuit.

6

T H E N O T E B O O K

The beat frequency will disappear at certain points as the deviation is in- - creased. These null points correspond to specific modulation indices, the first several being: 2.405, 5.520, 8.654, 11.792, and 14.931. Frequency deviation values at these null points are then cal- culated using the Bessel funct-ion as follows:

AF B = - f

Where: B = modulation index F = frequency deviation (kc) f z modulating frequency (kc)

After the first null is detected, the receiver is disconnected and the genera- tor signal is fed to the RF INPUT ter- minals on the Linear Detector. The De- tector is tuned to 350 kc IF frequency deviation to insure operation within the linear region. A peak-reading AC volt- meter connected to the DEMOD OUT terminals on the Detector will indicate the voltage output for the deviation cal- culated for the first null. The procedure is repeated for each null point, and the voltage output obtained for each calcu- lated frequency deviation is recorded. These voltages are then plotted against frequency deviation to produce a curve

I -1- .

Figure 4. Connections for percent AM measure- ment showing trapezoidal display.

whose straight-line pottion again in- dicates the linear limits of the Detector.

Application

The FM Linear Detector, as was pre- viously explained, has been designed especially for the measurement of FM frequency deviation. With the instru- ment calibrated as explained above, its frequency deviation versus voltage out- put coefficient is known. Measuring frequency deviation becomes merely a matter of multiplying the voltage output reading at the DEMOD OUT terminals of rhe Detector (produced as a result of the FM signal fed into the Detector) by the frequency deviation versus volt- age coefficient.

The Detector may also be used to measure the degree of amplitude modu- lation of a signal source. For this meas- urement, an external signal source is mixed with a signal produced by the Detector to provide a difference fre- quency of approximately 100 to 150 kc at the Detector’s IF OUTPUT terminals. This difference frequency is then dis- played on an oscilloscope as a trape- zoidal pattern, similar to the trapezoidal pattern shown in Figure 4. The lengths of the vertical sides of the pattern ( A and B) are then applied to the follow- ing equation to indicate the percentage of amplitude modulation.

A - B Percent AM = - A + B loo

Conclusion

The Type 208-A FM Linear Detector is a precision instrument capable of per- forming accurate FM and AM measure- ments in the engineering laboratory or on the production line. It will doubt- less become a valuable aid to those cus- tomers who have measurement problems of this nature.

BRC Film Gauge Used To Measure Aircraft Organic Finish Thickness

The Glenn 1. Martin Company Reports On a New Method Approved by The Navy

At the request of the Bureau of Aero- nautics, The Glenn L. Martin Company of Baltimore, Maryland recently pre- pared a report entitled, “A New Method for Measuring Aircraft Organic Finish Thickness,” which describes a new method employing the Boonton Radio Corporation Type 255-A Film Gauge.

The Film Gauge was introduced to The Martin Company by BRC as a means for measuring thickness of metallic plating finishes over nonferrous metal surfaces and its principle of opera- tion was later proposed as a means for measuring aircraft organic finish thick- ness. The Martin Company subsequently conducted a research program to test the suitability of the Film Gauge for

i/

this purpose.

Many finishes were tested by The Martin Company to find a correlation and mode of operation which would prove that practical measurements could be made with the Film Gauge. A cor- relation was found, preliminary calibra- tions were performed to confirm it, and a report proposing the new method was issued by The Martin Quality Division Laboratory. This report was later sub- mitted to the Bureau of Aerohautics who gave tentative approval of the method. The method was then evaluated by the Naval Air Material Center Labo- ratory. Results of this evaluation con- curred wih Martin’s and full acceptance of the method was published in a report

from Naval Air Material Center.

The report prepared by The Martin Company at the request of the Bureau of Aeronautics was prepared by Nor- man R. Keegan, Chemical-Physical Engi- neer. In his report Mr. Keegan describes in detail the methods used in determin- ing the accuracy of the Film Gauge for this specific application. Step-by-step procedures are given for operation of the Film Gauge for finish thickness in- spection, and special instructions are included for specific applications.

The Martin Report has been reprinted by Boonton Radio Corporation for dis- tribution to interested customers. Copies will be furnished upon request.

7

-*.

B O O N T O N R A D I O C O R P O R A T I O N T H E N O T E B O O K

EDITORS NOTE BRC Promotions Announced

The appointments of Mr. Frank G. Marble as Vice President and General Manager and Mr. Harry J. Lang as Sales Manager effective July 1st were an- nounced by Dr. George A. Downs- brough, President of Boonton Radio Corporation.

Mr. Marble, formerly Vice President- Sales at BRC, succeeds Dr. Downs- brough as General Manager. Dr. Downs-

Frank G. Marb le

brough will continue as President, Treasurer, and a Director.

Mr. Marble has been associated with Boonton Radio Corporation since 195 1. He served as Sales Manager until 1954 when he was appointed Vice President- Sales.

Prior to his associaion with BRC, Mr. Marble’s career covered a broad field of engineering experience. He held design and development posts for seven years with Philco Radio and Television Corporation and Western Electric’s elec- trical research division. Later, he served in engineering administrative positions with Bell Telephone Laboratories and Pratt and Whitney Aircraft Corporation. During the three-year period just pre- ceding his association with BRC, Mr. Marble served as Sales Manager for Kay Electric Company.

In 1934 Mr. Marble received his BS degree in Electrical Engineering from Mississippi State College. He earned his MS in Electrical Engineering from the Massachusetts Institute of Tech- nology in 1935.

Mr. Lang joined Boonton Radio Cor- poration in 1949 as a production engi- neer and successively served as project engineer, contracts engineer, and sales engineer. In 1954 he left BRC to study for his master’s in Business Administra- tion at the Harvard Business School.

Before returning to BRC, Mr. Lang served as sales engineer in charge of

sales for the newly-formed Industrial ~-,.-

Products Department of Airborne In-

Mr. Lang received his BS degree in Electrical Engineering from the Mas- sachusetts Institute of Technology in 1949. During his studies there, he was an engineering trainee at Western Elec- tric Co., New Jersey Bell Telephone Co., and Bell Telephone Laboratories.

From 1945 to 1946 he was a junior engineering officer with the U. S. Mer- chant Marine.

struments Laboratory. Ll

Harry J. Lang

ARLINGTON 74, Massachusefts DALLAS 9, Texas 10s ALTOS, California PITTSBURGH 36, Pennsylvania INSTRUMENT ASSOCIATES EARL LIPSCOMB ASSOCIATES VAN GROSS COMPANY H. E. RANSFORD COMPANY 1315 Massachusetts Avenue P. 0. Box 7084 1178 Los Altos Avenue 5400 Clairton 8oulevard Telephone: Mlssion 8-2922 Telephone: FLeetwood 7-1881 Telephone: WHitecliff 8-7266 Telephone: Tuxedo 4-3425 TWX: ARL MASS 253 TWX: DL 411

ROCHESTER 10, New York ATLANTA, Georgia DAYTON 9, Ohio E. A. OSSMANN & ASSOC., INC. 8lVlNS & CALDWELL CROSSLEY ASSO’S., INC. 830 Linden Avenue 3133 Maple Drive N.E. 53 Park Avenue Telephone: LUdlow 6-4940 Telephone: CEdar’3-7522 Telephone: OXmoor 3594

TWX: AT 987 TWX: DY 306 Telephone: Dlamond 0-3131 TWX: RO 189

E. A. OSSMANN & ASSOC., INC. 147 Front Street Telephone: Hlgh Point 3672 265 Church Street 842 Raymond Avenue Vestal, New York TWX: HIGH POINT NC 454 Telephone: UNiversity 5-2272 Telephone: Mldway 6-7881 Telephone: ENdicott 5-0296

10s ANGELES, California VAN GROSS COMPANY 21051 Costanso Street Post Office Box 425 Woodland Hills, California

NEW HAVEN 10, Connecticut HIGH POINT, North Carolina

ST. PAUL 14, Minnesota BINGHAMTON, New York 81VlNS 1923 North ’ CALDWELL Main Street INSTRUMENT ASSOCIATES CROSSLEY ASSO‘C., INC.

SYRACUSE, New York E. A. OSSMANN & ASSOC., INC. 308 Merritt Avenue Telephone: Howard 9-3825

TORONTO, Ontario, Canada

ORLANDO, Florida HOUSTON, Texas

8lVlNS & CALDWELL BOONTON, New Jersey EARL LIPSCOMB ASSOCIATES BOONTON RADIO CORPORATION ~ ; l ~ ; ~ ~ : 6 ~ ~ ~ k s o n 4-9303 1226 E. Colonial Drive lntervole Road Telephone: CHerry 1-1091 Telephone: DEerfield 4-3200 lNDlANAPOLI.5 20, Indiana TWX: BOONTON NJ 866 CROSSLEY ASSO‘S., INC.

5420 North College Avenue

Aiax, Ontario, Canodo Telephone: Aiox 118

(Toronto) EMpire 8-6866

CHICAGO 45, I l l inois CROSSLEY ASSO‘S., INC. 271 1 West Howard St. Telephone: SHeldroke 3-8500 TWX: CG 508

N E W JERSEY

Printed in U.S.A.

8